Control device, control method, and control program

Abstract

Provided is an IoT GW ( 10 ), including a learning unit ( 131 ) configured to create, for each IoT device connected to the IoT GW ( 10 ), a normal communication model ( 122 ) that has learned a normal communication pattern of the IoT device; and a determination unit ( 132 ) configured to: determine, for learning by the learning unit ( 131 ), whether to interrupt, finish, continue, and resume learning based on a load of a learning environment; and control learning processing by the learning unit ( 131 ) based on a result of determination.

The invention claimed is: 
     
       1. A control device, comprising:
 learning circuitry configured to create, for each IoT device connected to an IoT gateway, a normal communication model that has learned a normal communication pattern of the IoT device; and 
 determination circuitry configured to: 
 determine, for learning by the learning circuitry, whether to interrupt, finish, continue, and resume learning based on a load of a learning environment, the load on the learning environment corresponding to a percentage of usage of a CPU; and 
 control learning processing by the learning circuitry based on a result of determination, 
 wherein the determine to resume learning determines learning should be resumed when: 
 a period since an end of a previous learning is a multiple of a fixed value, 
 the load on the learning environment is smaller than a predetermined value, and 
 a number of devices being learned is smaller than a specific value. 
 
     
     
       2. The control device according to  claim 1 , further comprising:
 communication control circuitry configured to control communication by the IoT device by applying the normal communication model created by the learning circuitry. 
 
     
     
       3. The control device according to  claim 1 , wherein the determination circuitry comprises first determination circuitry configured to:
 determine, for each IoT device being learned, whether to interrupt or finish learning by the learning circuitry based on the load of the learning environment, a number of IoT devices to be learned, and a result of learning; 
 cause, when interruption of learning is determined, the learning circuitry to interrupt learning of the IoT device for which the interruption of learning has been determined; and 
 cause, when end of learning is determined, the learning circuitry to finish learning of the IoT device for which end of learning has been determined. 
 
     
     
       4. The control device according to  claim 3 , wherein the determination circuitry further comprises a second determination circuitry configured to:
 periodically determine, for each IoT device for which learning is interrupted, whether to allow continuation of interrupted learning based on the load of the learning environment, the number of IoT devices to be learned, and a length of an interruption period; and 
 cause, when continuation of learning is determined, the learning circuitry to continue learning of the IoT device for which continuation of learning has been determined. 
 
     
     
       5. The control device according to  claim 3 , wherein the determination circuitry further comprises a third determination circuitry configured to:
 periodically determine whether to allow resumption of learning based on the load of the learning environment and the number of IoT devices to be learned during a period since end of previous learning; and 
 cause the learning circuitry to resume learning when resumption of learning is determined. 
 
     
     
       6. A control method to be executed by a controller, the control method comprising:
 a learning process of creating, for each IoT device connected to an IoT gateway, a normal communication model that has learned a normal communication pattern of the IoT device; and 
 a determination process of: 
 determining, for learning by the learning process, whether to interrupt, finish, continue, and resume learning based on a load of a learning environment, the load on the learning environment corresponding to a percentage of usage of a CPU; and 
 controlling learning processing in the learning process based on a result of determination, 
 wherein the determining to resume learning determines learning should be resumed when: 
 a period since an end of a previous learning is a multiple of a fixed value, 
 the load on the learning environment is smaller than a predetermined value, and 
 a number of devices being learned is smaller than a specific value. 
 
     
     
       7. A non-transitory computer readable medium storing a control program for causing a computer to execute:
 a learning step of creating, for each IoT device connected to an IoT gateway, a normal communication model that has learned a normal communication pattern of the IoT device; and 
 a determination step of: 
 determining, for learning by the learning step, whether to interrupt, finish, continue, and resume learning based on a load of a learning environment, the load on the learning environment corresponding to a percentage of usage of a CPU; and 
 controlling learning processing in the learning step based on a result of determination, 
 wherein the determining to resume learning determines learning should be resumed when: 
 a period since an end of a previous learning is a multiple of a fixed value, 
 the load on the learning environment is smaller than a predetermined value, and 
 a number of devices being learned is smaller than a specific value. 
 
     
     
       8. The control device according to  claim 1 , wherein the determination circuitry is further configured to:
 determine whether the load on the learning environment is larger than the predetermined value; 
 determine whether the number of devices being learned is smaller than the specific value; and 
 determine whether a number of times of interruption and convergence rate are minimum values, 
 wherein when the load on the learning environment is determined to be larger than the predetermined value, the number of devices being learned is determined to be smaller than the specific value, and when the number of times of interruption and convergence rate are the minimum values, the learning is determined to be finished when the convergence rate is smaller than a fixed convergence value, and 
 wherein when the load on the learning environment is determined to be larger than the predetermined value, the number of devices being learned is determined to be smaller than the specific value, and when the number of times of interruption and convergence rate are the minimum values, the learning is determined to be interrupted when the convergence rate is not smaller than the fixed convergence value. 
 
     
     
       9. The control method according to  claim 6 , wherein the determining further comprises:
 determining whether the load on the learning environment is larger than the predetermined value; 
 determining whether the number of devices being learned is smaller than the specific value; and 
 determining whether a number of times of interruption and convergence rate are minimum values, 
 wherein when the load on the learning environment is determined to be larger than the predetermined value, the number of devices being learned is determined to be smaller than the specific value, and when the number of times of interruption and convergence rate are the minimum values, the learning is determined to be finished when the convergence rate is smaller than a fixed convergence value, and 
 wherein when the load on the learning environment is determined to be larger than the predetermined value, the number of devices being learned is determined to be smaller than the specific value, and when the number of times of interruption and convergence rate are the minimum values, the learning is determined to be interrupted when the convergence rate is not smaller than the fixed convergence value. 
 
     
     
       10. The non-transitory computer readable medium according to  claim 7 , wherein the determining further comprises:
 determining whether the load on the learning environment is larger than the predetermined value; 
 determining whether the number of devices being learned is smaller than the specific value; and 
 determining whether a number of times of interruption and convergence rate are minimum values, 
 wherein when the load on the learning environment is determined to be larger than the predetermined value, the number of devices being learned is determined to be smaller than the specific value, and when the number of times of interruption and convergence rate are the minimum values, the learning is determined to be finished when the convergence rate is smaller than a fixed convergence value, and 
 wherein when the load on the learning environment is determined to be larger than the predetermined value, the number of devices being learned is determined to be smaller than the specific value, and when the number of times of interruption and convergence rate are the minimum values, the learning is determined to be interrupted when the convergence rate is not smaller than the fixed convergence value.